| 67 | | // constants |
|---|
| 68 | | const IGNORE = '$1'; |
|---|
| 69 | | |
|---|
| 70 | | // validate parameters |
|---|
| 71 | | private $_script = ''; |
|---|
| 72 | | private $_encoding = 62; |
|---|
| 73 | | private $_fastDecode = true; |
|---|
| 74 | | private $_specialChars = false; |
|---|
| 75 | | |
|---|
| 76 | | private $LITERAL_ENCODING = array( |
|---|
| 77 | | 'None' => 0, |
|---|
| 78 | | 'Numeric' => 10, |
|---|
| 79 | | 'Normal' => 62, |
|---|
| 80 | | 'High ASCII' => 95 |
|---|
| 81 | | ); |
|---|
| 82 | | |
|---|
| 83 | | public function __construct($_script, $_encoding = 62, $_fastDecode = true, $_specialChars = false) |
|---|
| 84 | | { |
|---|
| 85 | | $this->_script = $_script . "\n"; |
|---|
| 86 | | if (array_key_exists($_encoding, $this->LITERAL_ENCODING)) |
|---|
| 87 | | $_encoding = $this->LITERAL_ENCODING[$_encoding]; |
|---|
| 88 | | $this->_encoding = min((int)$_encoding, 95); |
|---|
| 89 | | $this->_fastDecode = $_fastDecode; |
|---|
| 90 | | $this->_specialChars = $_specialChars; |
|---|
| 91 | | } |
|---|
| 92 | | |
|---|
| 93 | | public function pack() { |
|---|
| 94 | | $this->_addParser('_basicCompression'); |
|---|
| 95 | | if ($this->_specialChars) |
|---|
| 96 | | $this->_addParser('_encodeSpecialChars'); |
|---|
| 97 | | if ($this->_encoding) |
|---|
| 98 | | $this->_addParser('_encodeKeywords'); |
|---|
| 99 | | |
|---|
| 100 | | // go! |
|---|
| 101 | | return $this->_pack($this->_script); |
|---|
| 102 | | } |
|---|
| 103 | | |
|---|
| 104 | | // apply all parsing routines |
|---|
| 105 | | private function _pack($script) { |
|---|
| 106 | | for ($i = 0; isset($this->_parsers[$i]); $i++) { |
|---|
| 107 | | $script = call_user_func(array(&$this,$this->_parsers[$i]), $script); |
|---|
| 108 | | } |
|---|
| 109 | | return $script; |
|---|
| 110 | | } |
|---|
| 111 | | |
|---|
| 112 | | // keep a list of parsing functions, they'll be executed all at once |
|---|
| 113 | | private $_parsers = array(); |
|---|
| 114 | | private function _addParser($parser) { |
|---|
| 115 | | $this->_parsers[] = $parser; |
|---|
| 116 | | } |
|---|
| 117 | | |
|---|
| 118 | | // zero encoding - just removal of white space and comments |
|---|
| 119 | | private function _basicCompression($script) { |
|---|
| 120 | | $parser = new ParseMaster(); |
|---|
| 121 | | // make safe |
|---|
| 122 | | $parser->escapeChar = '\\'; |
|---|
| 123 | | // protect strings |
|---|
| 124 | | $parser->add('/\'[^\'\\n\\r]*\'/', self::IGNORE); |
|---|
| 125 | | $parser->add('/"[^"\\n\\r]*"/', self::IGNORE); |
|---|
| 126 | | // remove comments |
|---|
| 127 | | $parser->add('/\\/\\/[^\\n\\r]*[\\n\\r]/', ' '); |
|---|
| 128 | | $parser->add('/\\/\\*[^*]*\\*+([^\\/][^*]*\\*+)*\\//', ' '); |
|---|
| 129 | | // protect regular expressions |
|---|
| 130 | | $parser->add('/\\s+(\\/[^\\/\\n\\r\\*][^\\/\\n\\r]*\\/g?i?)/', '$2'); // IGNORE |
|---|
| 131 | | $parser->add('/[^\\w\\x24\\/\'"*)\\?:]\\/[^\\/\\n\\r\\*][^\\/\\n\\r]*\\/g?i?/', self::IGNORE); |
|---|
| 132 | | // remove: ;;; doSomething(); |
|---|
| 133 | | if ($this->_specialChars) $parser->add('/;;;[^\\n\\r]+[\\n\\r]/'); |
|---|
| 134 | | // remove redundant semi-colons |
|---|
| 135 | | $parser->add('/\\(;;\\)/', self::IGNORE); // protect for (;;) loops |
|---|
| 136 | | $parser->add('/;+\\s*([};])/', '$2'); |
|---|
| 137 | | // apply the above |
|---|
| 138 | | $script = $parser->exec($script); |
|---|
| 139 | | |
|---|
| 140 | | // remove white-space |
|---|
| 141 | | $parser->add('/(\\b|\\x24)\\s+(\\b|\\x24)/', '$2 $3'); |
|---|
| 142 | | $parser->add('/([+\\-])\\s+([+\\-])/', '$2 $3'); |
|---|
| 143 | | $parser->add('/\\s+/', ''); |
|---|
| 144 | | // done |
|---|
| 145 | | return $parser->exec($script); |
|---|
| 146 | | } |
|---|
| 147 | | |
|---|
| 148 | | private function _encodeSpecialChars($script) { |
|---|
| 149 | | $parser = new ParseMaster(); |
|---|
| 150 | | // replace: $name -> n, $$name -> na |
|---|
| 151 | | $parser->add('/((\\x24+)([a-zA-Z$_]+))(\\d*)/', |
|---|
| 152 | | array('fn' => '_replace_name') |
|---|
| 153 | | ); |
|---|
| 154 | | // replace: _name -> _0, double-underscore (__name) is ignored |
|---|
| 155 | | $regexp = '/\\b_[A-Za-z\\d]\\w*/'; |
|---|
| 156 | | // build the word list |
|---|
| 157 | | $keywords = $this->_analyze($script, $regexp, '_encodePrivate'); |
|---|
| 158 | | // quick ref |
|---|
| 159 | | $encoded = $keywords['encoded']; |
|---|
| 160 | | |
|---|
| 161 | | $parser->add($regexp, |
|---|
| 162 | | array( |
|---|
| 163 | | 'fn' => '_replace_encoded', |
|---|
| 164 | | 'data' => $encoded |
|---|
| 165 | | ) |
|---|
| 166 | | ); |
|---|
| 167 | | return $parser->exec($script); |
|---|
| 168 | | } |
|---|
| 169 | | |
|---|
| 170 | | private function _encodeKeywords($script) { |
|---|
| 171 | | // escape high-ascii values already in the script (i.e. in strings) |
|---|
| 172 | | if ($this->_encoding > 62) |
|---|
| 173 | | $script = $this->_escape95($script); |
|---|
| 174 | | // create the parser |
|---|
| 175 | | $parser = new ParseMaster(); |
|---|
| 176 | | $encode = $this->_getEncoder($this->_encoding); |
|---|
| 177 | | // for high-ascii, don't encode single character low-ascii |
|---|
| 178 | | $regexp = ($this->_encoding > 62) ? '/\\w\\w+/' : '/\\w+/'; |
|---|
| 179 | | // build the word list |
|---|
| 180 | | $keywords = $this->_analyze($script, $regexp, $encode); |
|---|
| 181 | | $encoded = $keywords['encoded']; |
|---|
| 182 | | |
|---|
| 183 | | // encode |
|---|
| 184 | | $parser->add($regexp, |
|---|
| 185 | | array( |
|---|
| 186 | | 'fn' => '_replace_encoded', |
|---|
| 187 | | 'data' => $encoded |
|---|
| 188 | | ) |
|---|
| 189 | | ); |
|---|
| 190 | | if (empty($script)) return $script; |
|---|
| 191 | | else { |
|---|
| 192 | | //$res = $parser->exec($script); |
|---|
| 193 | | //$res = $this->_bootStrap($res, $keywords); |
|---|
| 194 | | //return $res; |
|---|
| 195 | | return $this->_bootStrap($parser->exec($script), $keywords); |
|---|
| 196 | | } |
|---|
| 197 | | } |
|---|
| 198 | | |
|---|
| 199 | | private function _analyze($script, $regexp, $encode) { |
|---|
| 200 | | // analyse |
|---|
| 201 | | // retrieve all words in the script |
|---|
| 202 | | $all = array(); |
|---|
| 203 | | preg_match_all($regexp, $script, $all); |
|---|
| 204 | | $_sorted = array(); // list of words sorted by frequency |
|---|
| 205 | | $_encoded = array(); // dictionary of word->encoding |
|---|
| 206 | | $_protected = array(); // instances of "protected" words |
|---|
| 207 | | $all = $all[0]; // simulate the javascript comportement of global match |
|---|
| 208 | | if (!empty($all)) { |
|---|
| 209 | | $unsorted = array(); // same list, not sorted |
|---|
| 210 | | $protected = array(); // "protected" words (dictionary of word->"word") |
|---|
| 211 | | $value = array(); // dictionary of charCode->encoding (eg. 256->ff) |
|---|
| 212 | | $this->_count = array(); // word->count |
|---|
| 213 | | $i = count($all); $j = 0; //$word = null; |
|---|
| 214 | | // count the occurrences - used for sorting later |
|---|
| 215 | | do { |
|---|
| 216 | | --$i; |
|---|
| 217 | | $word = '$' . $all[$i]; |
|---|
| 218 | | if (!isset($this->_count[$word])) { |
|---|
| 219 | | $this->_count[$word] = 0; |
|---|
| 220 | | $unsorted[$j] = $word; |
|---|
| 221 | | // make a dictionary of all of the protected words in this script |
|---|
| 222 | | // these are words that might be mistaken for encoding |
|---|
| 223 | | //if (is_string($encode) && method_exists($this, $encode)) |
|---|
| 224 | | $values[$j] = call_user_func(array(&$this, $encode), $j); |
|---|
| 225 | | $protected['$' . $values[$j]] = $j++; |
|---|
| 226 | | } |
|---|
| 227 | | // increment the word counter |
|---|
| 228 | | $this->_count[$word]++; |
|---|
| 229 | | } while ($i > 0); |
|---|
| 230 | | // prepare to sort the word list, first we must protect |
|---|
| 231 | | // words that are also used as codes. we assign them a code |
|---|
| 232 | | // equivalent to the word itself. |
|---|
| 233 | | // e.g. if "do" falls within our encoding range |
|---|
| 234 | | // then we store keywords["do"] = "do"; |
|---|
| 235 | | // this avoids problems when decoding |
|---|
| 236 | | $i = count($unsorted); |
|---|
| 237 | | do { |
|---|
| 238 | | $word = $unsorted[--$i]; |
|---|
| 239 | | if (isset($protected[$word]) /*!= null*/) { |
|---|
| 240 | | $_sorted[$protected[$word]] = substr($word, 1); |
|---|
| 241 | | $_protected[$protected[$word]] = true; |
|---|
| 242 | | $this->_count[$word] = 0; |
|---|
| 243 | | } |
|---|
| 244 | | } while ($i); |
|---|
| 245 | | |
|---|
| 246 | | // sort the words by frequency |
|---|
| 247 | | // Note: the javascript and php version of sort can be different : |
|---|
| 248 | | // in php manual, usort : |
|---|
| 249 | | // " If two members compare as equal, |
|---|
| 250 | | // their order in the sorted array is undefined." |
|---|
| 251 | | // so the final packed script is different of the Dean's javascript version |
|---|
| 252 | | // but equivalent. |
|---|
| 253 | | // the ECMAscript standard does not guarantee this behaviour, |
|---|
| 254 | | // and thus not all browsers (e.g. Mozilla versions dating back to at |
|---|
| 255 | | // least 2003) respect this. |
|---|
| 256 | | usort($unsorted, array(&$this, '_sortWords')); |
|---|
| 257 | | $j = 0; |
|---|
| 258 | | // because there are "protected" words in the list |
|---|
| 259 | | // we must add the sorted words around them |
|---|
| 260 | | do { |
|---|
| 261 | | if (!isset($_sorted[$i])) |
|---|
| 262 | | $_sorted[$i] = substr($unsorted[$j++], 1); |
|---|
| 263 | | $_encoded[$_sorted[$i]] = $values[$i]; |
|---|
| 264 | | } while (++$i < count($unsorted)); |
|---|
| 265 | | } |
|---|
| 266 | | return array( |
|---|
| 267 | | 'sorted' => $_sorted, |
|---|
| 268 | | 'encoded' => $_encoded, |
|---|
| 269 | | 'protected' => $_protected); |
|---|
| 270 | | } |
|---|
| 271 | | |
|---|
| 272 | | private $_count = array(); |
|---|
| 273 | | private function _sortWords($match1, $match2) { |
|---|
| 274 | | return $this->_count[$match2] - $this->_count[$match1]; |
|---|
| 275 | | } |
|---|
| 276 | | |
|---|
| 277 | | // build the boot function used for loading and decoding |
|---|
| 278 | | private function _bootStrap($packed, $keywords) { |
|---|
| 279 | | $ENCODE = $this->_safeRegExp('$encode\\($count\\)'); |
|---|
| 280 | | |
|---|
| 281 | | // $packed: the packed script |
|---|
| 282 | | $packed = "'" . $this->_escape($packed) . "'"; |
|---|
| 283 | | |
|---|
| 284 | | // $ascii: base for encoding |
|---|
| 285 | | $ascii = min(count($keywords['sorted']), $this->_encoding); |
|---|
| 286 | | if ($ascii == 0) $ascii = 1; |
|---|
| 287 | | |
|---|
| 288 | | // $count: number of words contained in the script |
|---|
| 289 | | $count = count($keywords['sorted']); |
|---|
| 290 | | |
|---|
| 291 | | // $keywords: list of words contained in the script |
|---|
| 292 | | foreach ($keywords['protected'] as $i=>$value) { |
|---|
| 293 | | $keywords['sorted'][$i] = ''; |
|---|
| 294 | | } |
|---|
| 295 | | // convert from a string to an array |
|---|
| 296 | | ksort($keywords['sorted']); |
|---|
| 297 | | $keywords = "'" . implode('|',$keywords['sorted']) . "'.split('|')"; |
|---|
| 298 | | |
|---|
| 299 | | $encode = ($this->_encoding > 62) ? '_encode95' : $this->_getEncoder($ascii); |
|---|
| 300 | | $encode = $this->_getJSFunction($encode); |
|---|
| 301 | | $encode = preg_replace('/_encoding/','$ascii', $encode); |
|---|
| 302 | | $encode = preg_replace('/arguments\\.callee/','$encode', $encode); |
|---|
| 303 | | $inline = '\\$count' . ($ascii > 10 ? '.toString(\\$ascii)' : ''); |
|---|
| 304 | | |
|---|
| 305 | | // $decode: code snippet to speed up decoding |
|---|
| 306 | | if ($this->_fastDecode) { |
|---|
| 307 | | // create the decoder |
|---|
| 308 | | $decode = $this->_getJSFunction('_decodeBody'); |
|---|
| 309 | | if ($this->_encoding > 62) |
|---|
| 310 | | $decode = preg_replace('/\\\\w/', '[\\xa1-\\xff]', $decode); |
|---|
| 311 | | // perform the encoding inline for lower ascii values |
|---|
| 312 | | elseif ($ascii < 36) |
|---|
| 313 | | $decode = preg_replace($ENCODE, $inline, $decode); |
|---|
| 314 | | // special case: when $count==0 there are no keywords. I want to keep |
|---|
| 315 | | // the basic shape of the unpacking funcion so i'll frig the code... |
|---|
| 316 | | if ($count == 0) |
|---|
| 317 | | $decode = preg_replace($this->_safeRegExp('($count)\\s*=\\s*1'), '$1=0', $decode, 1); |
|---|
| 318 | | } |
|---|
| 319 | | |
|---|
| 320 | | // boot function |
|---|
| 321 | | $unpack = $this->_getJSFunction('_unpack'); |
|---|
| 322 | | if ($this->_fastDecode) { |
|---|
| 323 | | // insert the decoder |
|---|
| 324 | | $this->buffer = $decode; |
|---|
| 325 | | $unpack = preg_replace_callback('/\\{/', array(&$this, '_insertFastDecode'), $unpack, 1); |
|---|
| 326 | | } |
|---|
| 327 | | $unpack = preg_replace('/"/', "'", $unpack); |
|---|
| 328 | | if ($this->_encoding > 62) { // high-ascii |
|---|
| 329 | | // get rid of the word-boundaries for regexp matches |
|---|
| 330 | | $unpack = preg_replace('/\'\\\\\\\\b\'\s*\\+|\\+\s*\'\\\\\\\\b\'/', '', $unpack); |
|---|
| 331 | | } |
|---|
| 332 | | if ($ascii > 36 || $this->_encoding > 62 || $this->_fastDecode) { |
|---|
| 333 | | // insert the encode function |
|---|
| 334 | | $this->buffer = $encode; |
|---|
| 335 | | $unpack = preg_replace_callback('/\\{/', array(&$this, '_insertFastEncode'), $unpack, 1); |
|---|
| 336 | | } else { |
|---|
| 337 | | // perform the encoding inline |
|---|
| 338 | | $unpack = preg_replace($ENCODE, $inline, $unpack); |
|---|
| 339 | | } |
|---|
| 340 | | // pack the boot function too |
|---|
| 341 | | $unpackPacker = new JavaScriptPacker($unpack, 0, false, true); |
|---|
| 342 | | $unpack = $unpackPacker->pack(); |
|---|
| 343 | | |
|---|
| 344 | | // arguments |
|---|
| 345 | | $params = array($packed, $ascii, $count, $keywords); |
|---|
| 346 | | if ($this->_fastDecode) { |
|---|
| 347 | | $params[] = 0; |
|---|
| 348 | | $params[] = '{}'; |
|---|
| 349 | | } |
|---|
| 350 | | $params = implode(',', $params); |
|---|
| 351 | | |
|---|
| 352 | | // the whole thing |
|---|
| 353 | | return 'eval(' . $unpack . '(' . $params . "))\n"; |
|---|
| 354 | | } |
|---|
| 355 | | |
|---|
| 356 | | private $buffer; |
|---|
| 357 | | private function _insertFastDecode($match) { |
|---|
| 358 | | return '{' . $this->buffer . ';'; |
|---|
| 359 | | } |
|---|
| 360 | | private function _insertFastEncode($match) { |
|---|
| 361 | | return '{$encode=' . $this->buffer . ';'; |
|---|
| 362 | | } |
|---|
| 363 | | |
|---|
| 364 | | // mmm.. ..which one do i need ?? |
|---|
| 365 | | private function _getEncoder($ascii) { |
|---|
| 366 | | return $ascii > 10 ? $ascii > 36 ? $ascii > 62 ? |
|---|
| 367 | | '_encode95' : '_encode62' : '_encode36' : '_encode10'; |
|---|
| 368 | | } |
|---|
| 369 | | |
|---|
| 370 | | // zero encoding |
|---|
| 371 | | // characters: 0123456789 |
|---|
| 372 | | private function _encode10($charCode) { |
|---|
| 373 | | return $charCode; |
|---|
| 374 | | } |
|---|
| 375 | | |
|---|
| 376 | | // inherent base36 support |
|---|
| 377 | | // characters: 0123456789abcdefghijklmnopqrstuvwxyz |
|---|
| 378 | | private function _encode36($charCode) { |
|---|
| 379 | | return base_convert($charCode, 10, 36); |
|---|
| 380 | | } |
|---|
| 381 | | |
|---|
| 382 | | // hitch a ride on base36 and add the upper case alpha characters |
|---|
| 383 | | // characters: 0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ |
|---|
| 384 | | private function _encode62($charCode) { |
|---|
| 385 | | $res = ''; |
|---|
| 386 | | if ($charCode >= $this->_encoding) { |
|---|
| 387 | | $res = $this->_encode62((int)($charCode / $this->_encoding)); |
|---|
| 388 | | } |
|---|
| 389 | | $charCode = $charCode % $this->_encoding; |
|---|
| 390 | | |
|---|
| 391 | | if ($charCode > 35) |
|---|
| 392 | | return $res . chr($charCode + 29); |
|---|
| 393 | | else |
|---|
| 394 | | return $res . base_convert($charCode, 10, 36); |
|---|
| 395 | | } |
|---|
| 396 | | |
|---|
| 397 | | // use high-ascii values |
|---|
| 398 | | // characters: ¡¢£€¥Š§š©ª«¬Â®¯°±²³Žµ¶·ž¹º»ŒœŸ¿ÃÃÃÃÃà |
|---|
| | 67 | // constants |
|---|
| | 68 | const IGNORE = '$1'; |
|---|
| | 69 | |
|---|
| | 70 | // validate parameters |
|---|
| | 71 | private $_script = ''; |
|---|
| | 72 | private $_encoding = 62; |
|---|
| | 73 | private $_fastDecode = true; |
|---|
| | 74 | private $_specialChars = false; |
|---|
| | 75 | |
|---|
| | 76 | private $LITERAL_ENCODING = array( |
|---|
| | 77 | 'None' => 0, |
|---|
| | 78 | 'Numeric' => 10, |
|---|
| | 79 | 'Normal' => 62, |
|---|
| | 80 | 'High ASCII' => 95 |
|---|
| | 81 | ); |
|---|
| | 82 | |
|---|
| | 83 | public function __construct($_script, $_encoding = 62, $_fastDecode = true, $_specialChars = false) |
|---|
| | 84 | { |
|---|
| | 85 | $this->_script = $_script . "\n"; |
|---|
| | 86 | if (array_key_exists($_encoding, $this->LITERAL_ENCODING)) |
|---|
| | 87 | $_encoding = $this->LITERAL_ENCODING[$_encoding]; |
|---|
| | 88 | $this->_encoding = min((int)$_encoding, 95); |
|---|
| | 89 | $this->_fastDecode = $_fastDecode; |
|---|
| | 90 | $this->_specialChars = $_specialChars; |
|---|
| | 91 | } |
|---|
| | 92 | |
|---|
| | 93 | public function pack() { |
|---|
| | 94 | $this->_addParser('_basicCompression'); |
|---|
| | 95 | if ($this->_specialChars) |
|---|
| | 96 | $this->_addParser('_encodeSpecialChars'); |
|---|
| | 97 | if ($this->_encoding) |
|---|
| | 98 | $this->_addParser('_encodeKeywords'); |
|---|
| | 99 | |
|---|
| | 100 | // go! |
|---|
| | 101 | return $this->_pack($this->_script); |
|---|
| | 102 | } |
|---|
| | 103 | |
|---|
| | 104 | // apply all parsing routines |
|---|
| | 105 | private function _pack($script) { |
|---|
| | 106 | for ($i = 0; isset($this->_parsers[$i]); $i++) { |
|---|
| | 107 | $script = call_user_func(array(&$this,$this->_parsers[$i]), $script); |
|---|
| | 108 | } |
|---|
| | 109 | return $script; |
|---|
| | 110 | } |
|---|
| | 111 | |
|---|
| | 112 | // keep a list of parsing functions, they'll be executed all at once |
|---|
| | 113 | private $_parsers = array(); |
|---|
| | 114 | private function _addParser($parser) { |
|---|
| | 115 | $this->_parsers[] = $parser; |
|---|
| | 116 | } |
|---|
| | 117 | |
|---|
| | 118 | // zero encoding - just removal of white space and comments |
|---|
| | 119 | private function _basicCompression($script) { |
|---|
| | 120 | $parser = new ParseMaster(); |
|---|
| | 121 | // make safe |
|---|
| | 122 | $parser->escapeChar = '\\'; |
|---|
| | 123 | // protect strings |
|---|
| | 124 | $parser->add('/\'[^\'\\n\\r]*\'/', self::IGNORE); |
|---|
| | 125 | $parser->add('/"[^"\\n\\r]*"/', self::IGNORE); |
|---|
| | 126 | // remove comments |
|---|
| | 127 | $parser->add('/\\/\\/[^\\n\\r]*[\\n\\r]/', ' '); |
|---|
| | 128 | $parser->add('/\\/\\*[^*]*\\*+([^\\/][^*]*\\*+)*\\//', ' '); |
|---|
| | 129 | // protect regular expressions |
|---|
| | 130 | $parser->add('/\\s+(\\/[^\\/\\n\\r\\*][^\\/\\n\\r]*\\/g?i?)/', '$2'); // IGNORE |
|---|
| | 131 | $parser->add('/[^\\w\\x24\\/\'"*)\\?:]\\/[^\\/\\n\\r\\*][^\\/\\n\\r]*\\/g?i?/', self::IGNORE); |
|---|
| | 132 | // remove: ;;; doSomething(); |
|---|
| | 133 | if ($this->_specialChars) $parser->add('/;;;[^\\n\\r]+[\\n\\r]/'); |
|---|
| | 134 | // remove redundant semi-colons |
|---|
| | 135 | $parser->add('/\\(;;\\)/', self::IGNORE); // protect for (;;) loops |
|---|
| | 136 | $parser->add('/;+\\s*([};])/', '$2'); |
|---|
| | 137 | // apply the above |
|---|
| | 138 | $script = $parser->exec($script); |
|---|
| | 139 | |
|---|
| | 140 | // remove white-space |
|---|
| | 141 | $parser->add('/(\\b|\\x24)\\s+(\\b|\\x24)/', '$2 $3'); |
|---|
| | 142 | $parser->add('/([+\\-])\\s+([+\\-])/', '$2 $3'); |
|---|
| | 143 | $parser->add('/\\s+/', ''); |
|---|
| | 144 | // done |
|---|
| | 145 | return $parser->exec($script); |
|---|
| | 146 | } |
|---|
| | 147 | |
|---|
| | 148 | private function _encodeSpecialChars($script) { |
|---|
| | 149 | $parser = new ParseMaster(); |
|---|
| | 150 | // replace: $name -> n, $$name -> na |
|---|
| | 151 | $parser->add('/((\\x24+)([a-zA-Z$_]+))(\\d*)/', |
|---|
| | 152 | array('fn' => '_replace_name') |
|---|
| | 153 | ); |
|---|
| | 154 | // replace: _name -> _0, double-underscore (__name) is ignored |
|---|
| | 155 | $regexp = '/\\b_[A-Za-z\\d]\\w*/'; |
|---|
| | 156 | // build the word list |
|---|
| | 157 | $keywords = $this->_analyze($script, $regexp, '_encodePrivate'); |
|---|
| | 158 | // quick ref |
|---|
| | 159 | $encoded = $keywords['encoded']; |
|---|
| | 160 | |
|---|
| | 161 | $parser->add($regexp, |
|---|
| | 162 | array( |
|---|
| | 163 | 'fn' => '_replace_encoded', |
|---|
| | 164 | 'data' => $encoded |
|---|
| | 165 | ) |
|---|
| | 166 | ); |
|---|
| | 167 | return $parser->exec($script); |
|---|
| | 168 | } |
|---|
| | 169 | |
|---|
| | 170 | private function _encodeKeywords($script) { |
|---|
| | 171 | // escape high-ascii values already in the script (i.e. in strings) |
|---|
| | 172 | if ($this->_encoding > 62) |
|---|
| | 173 | $script = $this->_escape95($script); |
|---|
| | 174 | // create the parser |
|---|
| | 175 | $parser = new ParseMaster(); |
|---|
| | 176 | $encode = $this->_getEncoder($this->_encoding); |
|---|
| | 177 | // for high-ascii, don't encode single character low-ascii |
|---|
| | 178 | $regexp = ($this->_encoding > 62) ? '/\\w\\w+/' : '/\\w+/'; |
|---|
| | 179 | // build the word list |
|---|
| | 180 | $keywords = $this->_analyze($script, $regexp, $encode); |
|---|
| | 181 | $encoded = $keywords['encoded']; |
|---|
| | 182 | |
|---|
| | 183 | // encode |
|---|
| | 184 | $parser->add($regexp, |
|---|
| | 185 | array( |
|---|
| | 186 | 'fn' => '_replace_encoded', |
|---|
| | 187 | 'data' => $encoded |
|---|
| | 188 | ) |
|---|
| | 189 | ); |
|---|
| | 190 | if (empty($script)) return $script; |
|---|
| | 191 | else { |
|---|
| | 192 | //$res = $parser->exec($script); |
|---|
| | 193 | //$res = $this->_bootStrap($res, $keywords); |
|---|
| | 194 | //return $res; |
|---|
| | 195 | return $this->_bootStrap($parser->exec($script), $keywords); |
|---|
| | 196 | } |
|---|
| | 197 | } |
|---|
| | 198 | |
|---|
| | 199 | private function _analyze($script, $regexp, $encode) { |
|---|
| | 200 | // analyse |
|---|
| | 201 | // retrieve all words in the script |
|---|
| | 202 | $all = array(); |
|---|
| | 203 | preg_match_all($regexp, $script, $all); |
|---|
| | 204 | $_sorted = array(); // list of words sorted by frequency |
|---|
| | 205 | $_encoded = array(); // dictionary of word->encoding |
|---|
| | 206 | $_protected = array(); // instances of "protected" words |
|---|
| | 207 | $all = $all[0]; // simulate the javascript comportement of global match |
|---|
| | 208 | if (!empty($all)) { |
|---|
| | 209 | $unsorted = array(); // same list, not sorted |
|---|
| | 210 | $protected = array(); // "protected" words (dictionary of word->"word") |
|---|
| | 211 | $value = array(); // dictionary of charCode->encoding (eg. 256->ff) |
|---|
| | 212 | $this->_count = array(); // word->count |
|---|
| | 213 | $i = count($all); $j = 0; //$word = null; |
|---|
| | 214 | // count the occurrences - used for sorting later |
|---|
| | 215 | do { |
|---|
| | 216 | --$i; |
|---|
| | 217 | $word = '$' . $all[$i]; |
|---|
| | 218 | if (!isset($this->_count[$word])) { |
|---|
| | 219 | $this->_count[$word] = 0; |
|---|
| | 220 | $unsorted[$j] = $word; |
|---|
| | 221 | // make a dictionary of all of the protected words in this script |
|---|
| | 222 | // these are words that might be mistaken for encoding |
|---|
| | 223 | //if (is_string($encode) && method_exists($this, $encode)) |
|---|
| | 224 | $values[$j] = call_user_func(array(&$this, $encode), $j); |
|---|
| | 225 | $protected['$' . $values[$j]] = $j++; |
|---|
| | 226 | } |
|---|
| | 227 | // increment the word counter |
|---|
| | 228 | $this->_count[$word]++; |
|---|
| | 229 | } while ($i > 0); |
|---|
| | 230 | // prepare to sort the word list, first we must protect |
|---|
| | 231 | // words that are also used as codes. we assign them a code |
|---|
| | 232 | // equivalent to the word itself. |
|---|
| | 233 | // e.g. if "do" falls within our encoding range |
|---|
| | 234 | // then we store keywords["do"] = "do"; |
|---|
| | 235 | // this avoids problems when decoding |
|---|
| | 236 | $i = count($unsorted); |
|---|
| | 237 | do { |
|---|
| | 238 | $word = $unsorted[--$i]; |
|---|
| | 239 | if (isset($protected[$word]) /*!= null*/) { |
|---|
| | 240 | $_sorted[$protected[$word]] = substr($word, 1); |
|---|
| | 241 | $_protected[$protected[$word]] = true; |
|---|
| | 242 | $this->_count[$word] = 0; |
|---|
| | 243 | } |
|---|
| | 244 | } while ($i); |
|---|
| | 245 | |
|---|
| | 246 | // sort the words by frequency |
|---|
| | 247 | // Note: the javascript and php version of sort can be different : |
|---|
| | 248 | // in php manual, usort : |
|---|
| | 249 | // " If two members compare as equal, |
|---|
| | 250 | // their order in the sorted array is undefined." |
|---|
| | 251 | // so the final packed script is different of the Dean's javascript version |
|---|
| | 252 | // but equivalent. |
|---|
| | 253 | // the ECMAscript standard does not guarantee this behaviour, |
|---|
| | 254 | // and thus not all browsers (e.g. Mozilla versions dating back to at |
|---|
| | 255 | // least 2003) respect this. |
|---|
| | 256 | usort($unsorted, array(&$this, '_sortWords')); |
|---|
| | 257 | $j = 0; |
|---|
| | 258 | // because there are "protected" words in the list |
|---|
| | 259 | // we must add the sorted words around them |
|---|
| | 260 | do { |
|---|
| | 261 | if (!isset($_sorted[$i])) |
|---|
| | 262 | $_sorted[$i] = substr($unsorted[$j++], 1); |
|---|
| | 263 | $_encoded[$_sorted[$i]] = $values[$i]; |
|---|
| | 264 | } while (++$i < count($unsorted)); |
|---|
| | 265 | } |
|---|
| | 266 | return array( |
|---|
| | 267 | 'sorted' => $_sorted, |
|---|
| | 268 | 'encoded' => $_encoded, |
|---|
| | 269 | 'protected' => $_protected); |
|---|
| | 270 | } |
|---|
| | 271 | |
|---|
| | 272 | private $_count = array(); |
|---|
| | 273 | private function _sortWords($match1, $match2) { |
|---|
| | 274 | return $this->_count[$match2] - $this->_count[$match1]; |
|---|
| | 275 | } |
|---|
| | 276 | |
|---|
| | 277 | // build the boot function used for loading and decoding |
|---|
| | 278 | private function _bootStrap($packed, $keywords) { |
|---|
| | 279 | $ENCODE = $this->_safeRegExp('$encode\\($count\\)'); |
|---|
| | 280 | |
|---|
| | 281 | // $packed: the packed script |
|---|
| | 282 | $packed = "'" . $this->_escape($packed) . "'"; |
|---|
| | 283 | |
|---|
| | 284 | // $ascii: base for encoding |
|---|
| | 285 | $ascii = min(count($keywords['sorted']), $this->_encoding); |
|---|
| | 286 | if ($ascii == 0) $ascii = 1; |
|---|
| | 287 | |
|---|
| | 288 | // $count: number of words contained in the script |
|---|
| | 289 | $count = count($keywords['sorted']); |
|---|
| | 290 | |
|---|
| | 291 | // $keywords: list of words contained in the script |
|---|
| | 292 | foreach ($keywords['protected'] as $i=>$value) { |
|---|
| | 293 | $keywords['sorted'][$i] = ''; |
|---|
| | 294 | } |
|---|
| | 295 | // convert from a string to an array |
|---|
| | 296 | ksort($keywords['sorted']); |
|---|
| | 297 | $keywords = "'" . implode('|',$keywords['sorted']) . "'.split('|')"; |
|---|
| | 298 | |
|---|
| | 299 | $encode = ($this->_encoding > 62) ? '_encode95' : $this->_getEncoder($ascii); |
|---|
| | 300 | $encode = $this->_getJSFunction($encode); |
|---|
| | 301 | $encode = preg_replace('/_encoding/','$ascii', $encode); |
|---|
| | 302 | $encode = preg_replace('/arguments\\.callee/','$encode', $encode); |
|---|
| | 303 | $inline = '\\$count' . ($ascii > 10 ? '.toString(\\$ascii)' : ''); |
|---|
| | 304 | |
|---|
| | 305 | // $decode: code snippet to speed up decoding |
|---|
| | 306 | if ($this->_fastDecode) { |
|---|
| | 307 | // create the decoder |
|---|
| | 308 | $decode = $this->_getJSFunction('_decodeBody'); |
|---|
| | 309 | if ($this->_encoding > 62) |
|---|
| | 310 | $decode = preg_replace('/\\\\w/', '[\\xa1-\\xff]', $decode); |
|---|
| | 311 | // perform the encoding inline for lower ascii values |
|---|
| | 312 | elseif ($ascii < 36) |
|---|
| | 313 | $decode = preg_replace($ENCODE, $inline, $decode); |
|---|
| | 314 | // special case: when $count==0 there are no keywords. I want to keep |
|---|
| | 315 | // the basic shape of the unpacking funcion so i'll frig the code... |
|---|
| | 316 | if ($count == 0) |
|---|
| | 317 | $decode = preg_replace($this->_safeRegExp('($count)\\s*=\\s*1'), '$1=0', $decode, 1); |
|---|
| | 318 | } |
|---|
| | 319 | |
|---|
| | 320 | // boot function |
|---|
| | 321 | $unpack = $this->_getJSFunction('_unpack'); |
|---|
| | 322 | if ($this->_fastDecode) { |
|---|
| | 323 | // insert the decoder |
|---|
| | 324 | $this->buffer = $decode; |
|---|
| | 325 | $unpack = preg_replace_callback('/\\{/', array(&$this, '_insertFastDecode'), $unpack, 1); |
|---|
| | 326 | } |
|---|
| | 327 | $unpack = preg_replace('/"/', "'", $unpack); |
|---|
| | 328 | if ($this->_encoding > 62) { // high-ascii |
|---|
| | 329 | // get rid of the word-boundaries for regexp matches |
|---|
| | 330 | $unpack = preg_replace('/\'\\\\\\\\b\'\s*\\+|\\+\s*\'\\\\\\\\b\'/', '', $unpack); |
|---|
| | 331 | } |
|---|
| | 332 | if ($ascii > 36 || $this->_encoding > 62 || $this->_fastDecode) { |
|---|
| | 333 | // insert the encode function |
|---|
| | 334 | $this->buffer = $encode; |
|---|
| | 335 | $unpack = preg_replace_callback('/\\{/', array(&$this, '_insertFastEncode'), $unpack, 1); |
|---|
| | 336 | } else { |
|---|
| | 337 | // perform the encoding inline |
|---|
| | 338 | $unpack = preg_replace($ENCODE, $inline, $unpack); |
|---|
| | 339 | } |
|---|
| | 340 | // pack the boot function too |
|---|
| | 341 | $unpackPacker = new JavaScriptPacker($unpack, 0, false, true); |
|---|
| | 342 | $unpack = $unpackPacker->pack(); |
|---|
| | 343 | |
|---|
| | 344 | // arguments |
|---|
| | 345 | $params = array($packed, $ascii, $count, $keywords); |
|---|
| | 346 | if ($this->_fastDecode) { |
|---|
| | 347 | $params[] = 0; |
|---|
| | 348 | $params[] = '{}'; |
|---|
| | 349 | } |
|---|
| | 350 | $params = implode(',', $params); |
|---|
| | 351 | |
|---|
| | 352 | // the whole thing |
|---|
| | 353 | return 'eval(' . $unpack . '(' . $params . "))\n"; |
|---|
| | 354 | } |
|---|
| | 355 | |
|---|
| | 356 | private $buffer; |
|---|
| | 357 | private function _insertFastDecode($match) { |
|---|
| | 358 | return '{' . $this->buffer . ';'; |
|---|
| | 359 | } |
|---|
| | 360 | private function _insertFastEncode($match) { |
|---|
| | 361 | return '{$encode=' . $this->buffer . ';'; |
|---|
| | 362 | } |
|---|
| | 363 | |
|---|
| | 364 | // mmm.. ..which one do i need ?? |
|---|
| | 365 | private function _getEncoder($ascii) { |
|---|
| | 366 | return $ascii > 10 ? $ascii > 36 ? $ascii > 62 ? |
|---|
| | 367 | '_encode95' : '_encode62' : '_encode36' : '_encode10'; |
|---|
| | 368 | } |
|---|
| | 369 | |
|---|
| | 370 | // zero encoding |
|---|
| | 371 | // characters: 0123456789 |
|---|
| | 372 | private function _encode10($charCode) { |
|---|
| | 373 | return $charCode; |
|---|
| | 374 | } |
|---|
| | 375 | |
|---|
| | 376 | // inherent base36 support |
|---|
| | 377 | // characters: 0123456789abcdefghijklmnopqrstuvwxyz |
|---|
| | 378 | private function _encode36($charCode) { |
|---|
| | 379 | return base_convert($charCode, 10, 36); |
|---|
| | 380 | } |
|---|
| | 381 | |
|---|
| | 382 | // hitch a ride on base36 and add the upper case alpha characters |
|---|
| | 383 | // characters: 0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ |
|---|
| | 384 | private function _encode62($charCode) { |
|---|
| | 385 | $res = ''; |
|---|
| | 386 | if ($charCode >= $this->_encoding) { |
|---|
| | 387 | $res = $this->_encode62((int)($charCode / $this->_encoding)); |
|---|
| | 388 | } |
|---|
| | 389 | $charCode = $charCode % $this->_encoding; |
|---|
| | 390 | |
|---|
| | 391 | if ($charCode > 35) |
|---|
| | 392 | return $res . chr($charCode + 29); |
|---|
| | 393 | else |
|---|
| | 394 | return $res . base_convert($charCode, 10, 36); |
|---|
| | 395 | } |
|---|
| | 396 | |
|---|
| | 397 | // use high-ascii values |
|---|
| | 398 | // characters: ¡¢£€¥Š§š©ª«¬Â®¯°±²³Žµ¶·ž¹º»ŒœŸ¿ÃÃÃÃÃà |
|---|
